The expression dancing landscape is a good allegory of this characteristic of our business environment. The challenge of the strategy in the dancing landscape converts into two wicked problems: How to measure and find the strategy performance peaks? And how to balance your focus between exploiting the peaks already conquered and exploring new ones?
Imagine a landscape consisting of a vast grid with billions of squares where each square contains a business strategy. Potentially this grid covers all possible strategies of all possible businesses. The picture below describes a part of this landscape, e.g. your current industry and your and your competitors’ strategies.
The squares close to each other are strategies differing only a little from each other and squares distant from each other are totally different strategies, usually plans of totally different businesses. The height of each square – marked also with changing colors like in topological maps – describes the fitness of each business plan, for example the profitability (ROI, IRR or similar). The fitness dimension – the height – is the reason for these grids being called fitness landscapes. Mountain tops are good strategies, valleys are bad strategies. Most of the grids are strategies that nobody would even consider or totally impossible in the real world or pure gibberish.
Continuous fluctuation of the business environment
The challenge in estimating the fitness (profitability) of each strategy, is the continuous fluctuation of the landscape. The fitness of the strategy changes over time. Yesterday’s profitable strategies can overnight turn into an unprofitable endeavor. It is not because of the weakness of the strategy as such, but because of the changes that take place in the landscape – what our competitors, customers, employees, government etc. do. The landscape is a canvas under which different actors move causing the whole landscape to change continuously. The expression dancing landscape is a good allegory of this characteristic of our business environment.
Fitness landscape is a geometrically motivated tool to visualize and evaluate how a strategy may evolve over time. Fitness landscapes originated from evolutionary biology in the 1930s as a tool to map from a set of genotypes to fitness. The orthogonal projection of the genotype is the property called fitness reflecting the capability of the genotype to survive and reproduce. The genotypes were countable and neighboring in such a way that genotypes next to each other can mutate from one to another. The analogy between the genotype and strategies is self-evident – both determine in matters of life and death.
Winning strategies are temporary
Our business environment has typically many peaks occupied by the strategies of different companies. Different strategies can survive in the same industry. Some do better than others, some worse. Only strategies that exceed a critical level of fitness (usually a certain profitability threshold) survive in the marketplace. But no one can count on the continuity of the success. Past success is not a guarantee of future success, especially in a dancing landscape.
The key characteristic of such a business environment is its unpredictability. Reliable forecasting of the changes in the landscape is impossible. As long as our stakeholders consist of human beings with free will, anything is possible and nothing is impossible. As a consequence, there is no such thing as a winning strategy ex ante, only ex post, after we have seen how others react to it and how the world changes. We all are familiar with famous business books telling about winning strategies: In Search of Excellence, Good to Great, What Really Works, Built to Last etc. The only problem is that the world is full of companies who have followed the same strategies and failed badly as Philip Rosenzweig in his book “Halo Effect” unequivocally showed.
Although changes in the business landscape cannot be controlled, an approximate map of the landscape can be drawn out by constantly surveying it. There are no readymade maps, but fast paced, parallel trials are the best approach in identifying contours in the landscape and in shaping and conquering new peaks. We can see a foggy picture of the current landscape but we cannot trust those contours and peaks to stay there for any longer period. Still the machinery of strategic planning is often based on analyzing this picture, forecasting its development – sometimes over a period of one year and in extreme cases of several years – and developing strategies to conquer the assumed peaks of the landscape. But like Field Marshall Helmuth von Moltke the Elder said, “No plan survives the contact of the enemy”. Planning as such is an exercise in futility, but still we plan – and especially the military does. The main reason for this is to recognize and understand the changes once they come, to separate signals from noise. Plans are there to help us living with the evitable changes.
Key learnings of strategy development in an unpredictable environment
We can also learn from the dancing landscape to build our plans on, to plan for the unexpected. Regarding planning and managing there are six key differences between the traditional stationary view of the business landscape (Stationary landscape, SL) and the Dancing Landscape (DL):
- SL is based on ability to forecast long term developments, whereas DL requires accepting alternative development paths and using only short term forecasting to lock the situation for fast and determined actions.
- SL favors a binding and irrevocable nature of choices to show commitment to the strategy, DL favors flexibility, variety of choices and real options.
- SL appreciates sustainable and long-term competitive advantage, DL favors capabilities that enable constant renewal.
- SL requires clear and shared intent throughout the whole organization and strong commitment from everybody, DL appreciates fast learning, variation and adaptation near the customer.
- A strategy built on SL has usually a few major development projects with clearly expressed predetermined targets (often must-win-battles), DL favors several parallel fast trials, failing fast and copying successful trials.
- SL culture and modus operandi is planning based on centralized decision making, DL builds on responsiveness, listening and discussing, giving mandate to act to those close the action.
This distinction is on purpose a black and white interpretation to accentuate the differences. The real world of strategies and management is a combination of all these. Perhaps strategy should be developed more on the terms of the Dancing Landscape and the implementation on the terms of the Stationary Landscape.
Two problems: search and balance
The challenge of the strategy in the dancing landscape converts ultimately into two wicked problems: 1. How to find the peaks? and 2. How to balance your focus between exploiting the peaks already conquered and exploring new ones?
The first problem, also known as a search problem, focuses on analyzing the landscape to find the most efficient approach to maximize the fitness of our solution, in our case, what is the best strategy for this industry. The second problem starts from the fact that as the peaks of the landscape are changing, we cannot only focus on exploiting a high one – in our case implementing a profitable strategy – once we have found it, but continue searching new peaks.
Analogies as a source of new solutions in business
These two questions – being fundamental issues in strategy and management – are also paramount in many other fields and sciences. This opens us an opportunity of piggy bagging on approaches and solutions developed in other areas than business, e.g. biology, physics, mathematics and especially computer science. These sciences have studied and developed their own approaches and solutions to our two fundamental questions. Approaches and solutions that are often in disguise and require substantial interpretation and adaptation, but ultimately handle the very same issue as we do.
One fascinating example is a conclusion developed in the complexity science where studies in multidimensional landscapes have shown that with sufficient representational diversity, any problem can be solved.
In our case this conclusion says that it is possible to find the best strategy in a multidimensional landscape by having enough different views on the issue, i.e. being able to represent the problem in enough different ways. A practical application of this research conclusion is that when developing strategies for a complex environment it is good to have enough people with totally different backgrounds and views to analyze the situation and to develop the strategy. Diversity in the approach should increase in relation to the diversity of the problem.
Mathematically the previous example would actually have been a little more complicated to express, but in a very near future we have artificial intelligence assistants which can surpass all these extra complications and support us navigating and optimizing our strategies in a dancing landscape. The future of strategies and strategy development will not be what it used to be.
Escrito por Martti Malmivirta, EERA Oy Executive Chairman.
References:
Sewall Wright (1932). «The roles of mutation, inbreeding, crossbreeding and selection in evolution» in D.F. Jones (ed.) Proc. Of the Sixth International Congress on Genetics pp. 356–366.
Hendrik Richter, Andres P. Engelbrecht (ed.) (2014) Recent Advances in the Theory and Application of Fitness Landscapes, Springer Verlag, Berlin Heidelberg
Tom Peters and Robert H. Waterman (1982) In Search of Excellence, Harper & Row, New York
Jim Collins (2001) Good to Great: Why Some Companies Make the Leap…and Others Don’t. Harper Business, New York
Jim Collins and Jerry Porras (2002), Built to Last, Successful Habit of Visionary Companies, HarperCollins, New York
William Joyce, Nitin Nohria and Bruce Roberson (2004) What Really Works: The 4+2 Formula for Sustained Business Success, Harper Business, New York
Philip Rosenzweig (2007) The Halo Effect: . . . and the Eight Other Business Delusions That Deceive Managers, Free Press, New York
Lu Hong and Scott Page (2004) Groups of diverse problem solvers can outperform groups of high-ability problem solvers, Proceedings of the National Academy of Sciencess 101 (46), pp 16385-89.
